Many bacterial treatments are directed to bacteria in a planktonic state. However, bacterial pathologies include bacteria in a biofilm state. For example, Porphyromonas gingivalis is considered to be the major causative agent of chronic periodontal disease. Tissue damage associated with the disease is caused by a dysregulated host immune response to P. gingivalis growing as a part of a polymicrobial bacterial biofilm on the surface of the tooth. Bacterial biofilms are ubiquitous in nature and are defined as matrix-enclosed bacterial populations adherent to each other and/or to surfaces or interfaces (1). These sessile bacterial cells adhering to and growing on a surface as a mature biofilm are able to survive in hostile environments which can include the presence of antimicrobial agents, shear forces and nutrient deprivation.
The Centers for Disease Control and Prevention estimate that 65% of human bacterial infections involve biofilms. Biofilms often complicate treatment of chronic infections by protecting bacteria from the immune system, decreasing antibiotic efficacy and dispersing planktonic cells to distant sites that can aid reinfection (2,3). Dental plaque is a classic example of a bacterial biofilm where a high diversity of species form a heterogeneous polymicrobial biofilm growing on the surface of the tooth. The surface of the tooth is a unique microbial habitat as it is the only hard, permanent, non-shedding surface in the human body. This allows the accretion of a substantial bacterial biofilm over a lengthy time period as opposed to mucosal surfaces where epithelial cell shedding limits development of the biofilm. Therefore, the changes to the P. gingivalis proteome that occur between the planktonic and biofilm states are important to our understanding of the progression of chronic periodontal disease.
P. gingivalis has been classified into two broad strain groups with strains including W50 and W83 being described as invasive in animal models whilst strains including 381 and ATCC 33277 are described as non-invasive (4,5). Griffen et al. (6) found that W83/W50-like strains were more associated with human periodontal disease than other P. gingivalis strains, including 381-like strains, whilst Cutler et al. (7) demonstrated that invasive strains of P. gingivalis were more resistant to phagocytosis than non-invasive strains. Comparison of the sequenced P. gingivalis W83 strain to the type strain ATCC 33277 indicated that 7% of genes were absent or highly divergent in strain 33277 indicating that there are considerable differences between the strains (8). Interestingly P. gingivalis strain W50 forms biofilms only poorly under most circumstances compared to strain 33277 which readily forms biofilms (9). As a consequence of this relatively few studies have been conducted on biofilm formation by P. gingivalis W50.
Quantitative proteomic studies have been employed to determine proteome changes of human bacterial pathogens such as Pseudomonas aeruginosa, Escherichia coli and Streptococcus mutans from the planktonic to biofilm state using 2D gel electrophoresis approaches, where protein ratios are calculated on the basis of gel staining intensity (10-12). An alternative is to use stable isotope labelling techniques such as ICAT, iTRAQ or heavy water (H218O) with MS quantification (13). The basis for H218O labelling is that during protein hydrolysis endopeptidases such as trypsin have been demonstrated to incorporate two 18O atoms into the C-termini of the resulting peptides (14,15). In addition to use in the determination of relative protein abundances (16-19), 18O labelling in proteomics has also been used for the identification of the protein C-terminus, identification of N-linked glycosylation after enzymatic removal of the glycan, simplification of MS/MS data interpretation and more recently for validation of phosphorylation sites (20-23). The 16O/18O proteolytic labelling method for measuring relative protein abundance involves digesting one sample in H216O and the other sample in H218O. The digests are then combined prior to analysis by LC MS/MS. Peptides eluting from the LC column can be quantified by measuring the relative signal intensities of the peptide ion pairs in the MS mode. The incorporation of two 18O atoms into the C-terminus of digested peptides by trypsin results in a mass shift of +4 m/z allowing the identification of the isotope pairs.
Due to the complexity of the proteome, prefractionation steps are advantageous for increasing the number of peptide and protein identifications. Most prefractionation steps involve a 2D LC approach at the peptide level after in-solution digestion (24,25). However due to potential sample loss during the initial dehydration steps of the protein solution, SDS PAGE prefractionation at the protein level followed by 16O/18O labelling during in gel digestion has also been carried out successfully, (26-29). The 16O/18O proteolytic labelling is a highly specific and versatile methodology but few validation studies on a large scale have been performed (30). An excellent validation study was carried out by Qian et al (18) who labelled two similar aliquots of serum proteins in a 1:1 ratio and obtained an average ratio of 1.02±0.23 from 891 peptides. A more recent study by Lane et al (26) further demonstrated the feasibility of the 16O/18O method using a reverse labelling strategy to determine the relative abundance of 17 cytochrome P450 proteins between control and cytochrome P450 inducers treated mice that are grafted with human tumours.